Hacker News new | past | comments | ask | show | jobs | submit login
Products I Wish Existed (eladgil.com)
798 points by eladgil on Jan 6, 2020 | hide | past | favorite | 419 comments

+100 "I would love to see the following analysis: A map of repetitious tasks, spreadsheets, and manual data extraction by function in the Fortune 500. Budget breakdown of current software spend, by function, by line, in the Fortune 500. A view of what Accenture, CapGemini, and Deloitte keep building over and over for large enterprises. Undoubtedly a subset of these custom consulting projects can be turned into SaaS software. A tougher analysis to do is to ask what internal software projects various tech companies are working on. If you can get the list from 3-4 companies, you will undoubtedly see a few internal tool or product examples that should be built as a SaaS product for everyone."

* I'd love to see that as well

* At the same time:

- A LOT of companies are uncomfortable making some of this data SaaS / external. I'm not saying they're right, but it's there

- A LOT of companies are unique, or strongly believe they're unique, and that they need bespoke / highly customized software. I'm not saying they're right... ;)

- I'm a techie, so it's taken me a decade to realize that implementing a new backoffice suite by a consulting company is one third or less technical / IT project; and two thirds or more business process / transformation project. And the actual success / failure of these projects is almost entirely guided by the business transformation part.

And when a company views ERP as a technical/IT project, it is most likely to not end well (small scale example: Sally in accounting refuses to change and insists that the new fancy software, on-prem or SaaS, do things the way she's done them for 2 decades in Excel, because that's "how their company does it"... no matter how inefficient or outdated they are. If you're seen as an IT project, it's your duty to accommodate Sally and now you're building the same thing over and over, compromising any value in your new ERP, SaaS or otherwise. If you're seen as business transformation project, you have some chance to make a difference and make things better...]

- Therefore SaaS would reduce FAR less effort in ERP than many techies (including myself) feel. Basically, you'd still have a consulting company come in to (try) to do business transformation so your internal processes actually accept and match the SaaS processes; build interfaces; and then maintain and support.

Disclosure/example: I'm in consulting part of IBM, and while I'm currently on a legacy ERP project, my colleagues in "Cloud/SaaS" (i.e. "let somebody else host it for you") side are no less busy than before. Mind you, they are on average providing higher value services, so that's a plus :)

>implementing a new backoffice suite by a consulting company is one third or less technical / IT project; and two thirds or more business process / transformation project. And the actual success / failure of these projects is almost entirely guided by the business transformation part.

This 1000%. There's so much institutional friction with these projects for so many reasons; some of them political, some of them interpersonal, some of them just because someone doesn't want to. So many of these projects fail because people just don't want to work towards the goal, or simply do as little as possible and things move slower than molasses.

People don't want or care to change their processes. They do things that way because they've always done things that way. They see tools not as a way to get more done or be efficient, but as something that's "moving their cheese" and they don't like it. It doesn't matter how overworked they are by remedial tasks that could be automated, until you show them everything finished, end-to-end, working flawlessly, they won't be interested in the slightest.

The sad thing is that if there were some strong executive leadership behind these projects and they saw it as a positive thing, there could be better traction and better results. Instead, most execs look at these projects as mundane details and hand them off to whatever sad middle manager is going to take the heat for the project going south. The result is no teeth behind the initiative which even further exacerbates the issue of people not caring or lifting a finger.

> They see tools not as a way to get more done or be efficient, but as something that's "moving their cheese" and they don't like it.

In all fairness, this perception of tools is often entirely rational. Changing workflows incurs a high cost, and any new tool must supply a benefit in excess of that cost. If it doesn't, then opposing the new tool can be the correct stance. And in my experience, at least half of the time, the new tools do not provide sufficient benefit.

Exactly. There's endless tooling out there, and you can contort software to whatever process.

Some of tools make claims that, if trivial to implement, would provide huge huge benefits. Unfortunately, there's a cost to learning a new tool, and then to implementing its use. Since you don't know the tool yet, you can only make educated guesses as to whether it will be easy or hard to learn, easy or hard to implement, make your process better or worse, help sell your product or make you more efficient, etc. You can't know for sure (unless it seems so much like something you already DO understand, and in that case, why would you use someone else's tool)? And that's only for yourself.

I'm not saying you shouldn't try new tools. To the contrary. It's just hard to know which ones are the right ones for your organization. Hence cargo-culting of tools that may or may not be a step backwards for your organization, or have all kinds of hidden-costs that are difficult or even impossible to foresee.

They see tools not as a way to get more done or be efficient, but as something that's "moving their cheese" and they don't like it. It doesn't matter how overworked they are by remedial tasks that could be automated, until you show them everything finished, end-to-end, working flawlessly, they won't be interested in the slightest.

Will they be paid more for the increase in productivity?

I looked into content rating systems one time. Specifically PICS[1] As it gained some attention in some big media outlets where authoritative sounding journalists full on attacked it. Some attempted to portray it as a filter. At first I didn't understand why the persistent negativity but eventually I got it when one of them wrote how he acquired his skills though years of article writing and that the public scrutiny was an insult to his profession.

I then began to notice this pattern in other people. If you give them efficient software they wont be able to bullshit their way out of any argument. Productivity is the enemy. Everything possible must be done to prevent the people above from figuring out what is really going on.

One very simple example I ran into recently. A new and improved app was made but it had exactly the same flaws as the previous one. An amount of time is assigned to tasks but if you add up the tasks for a day they greatly exceed the real world number of minutes worked. Then when complaints come in they simply assign extra minutes to the task pushing the total amount of work further over 8 hours. The number of characters for feedback was also reduced. When I jokingly asked about it (knowing how their game works) they pretended it was to hard to divide the work over the day. You cant just say a 70 min task must be done in 50! All hell would break loose! It is then left up to the assignee to figure out how to cut the corners. The managers carefully monitor the feedback, if anyone finds out corners are cut the employee is called in to explain for it. Eventually the smart ones figure out how to cut corners without anyone noticing it.

The entire show depends on the clients inability to add up all the minutes.

[1] - https://www.w3.org/PICS/services-960303.html

> companies are uncomfortable making some of this data SaaS / external

SaaS can be on-premises.

> Sally in accounting refuses to change

The tools need to adapt to the user, the user shouldn't need to adapt to the tool. That said, if there are obvious efficiency gains (e.g. you can stop typing "COMPLETED" three hundred times a day and just click a button that fills the field for you; even better, the process marks it 'completed' once all the human verification has taken place) and the user is stubborn, well, maybe the tool should offer the efficient solution, allow the inefficient behavior, and log metrics on the whole thing. Management might take interest.

>>SaaS can be on-premises.

True; the distinction between SaaS and COTS becomes moot in practical terms then, and all of my comment still applies.

>>The tools need to adapt to the user, the user shouldn't need to adapt to the tool.

Again, that's:

- True for an IT project whose goal is to support whatever the user is currently doing (and what many of us are trained to do / think in current IT paradigm).

- Not True for a business transformation project whose goal is to change business processes for the better - and, basically incidentally, deliver a COTS ERP to support them through an IT portion of the project.

We are not talking about where the button is or how many times it takes to click or what the icon looks like. We are talking fundamental business processes and workflow, which are core to how a business operates internally.

For business processes which are company's core business, presumably they have them figured out. That's great, and IT should support them.

For back-office ERP, this is not your company's core business; you're probably not a special snowflake; and the market's average/best practices are likely miles ahead of your legacy over-complicated business processes. YOU need to change if you want to reap the benefits of shiny expensive new software.

There is extremely limited point in implementing a new back-office ERP (HR, Financials, CRM, EPM, etc) if you're not willing to approach it as BT project; radically change how you do things; and expect a shiny new tool to make you more competitive automagically without your processes and maybe even culture/habits changing.

My fundamental point remains: what the big consulting companies do with ERP / backoffice implementation (when done right) are primarily BT projects, and incidentally involve IT implementations; and treating them as IT projects, whether that's Sally in accounting refusing to change or Bob the developer accommodating Sally thinking it's the right thing to do, will cause them to fail. (generalizations and oversimplifications abound in above statement; but it really takes a very very large and repeated application of a mental sledgehammer to dislodge some universally-good but specifically-counterproductive ideas and habits from the ERP space)

Quite a few years ago, I was at a very small company and we made the decision to move from hosting our own Microsoft Exchange to Gmail--partly for cost reasons and partly because we had had a couple of serious mail outages.

A couple of people at the time were really unhappy about the change because there were some differences, at least at the time, in the ability to create nested folders/labels like they were accustomed to doing in Exchange. Basically, Gmail messed up the organizational scheme they were accustomed to for client and contracts.

They were basically told to deal with it but the point is that even when going to some standard SaaS makes sense, people are very resistant to making changes in their standard tools.

In my first startup, I switched our email from an ISP hosted offering to an self-hosted system. We were ramen poor, and this was projected to save us around $1K/month. The replacement system was postfix with dovecot, worked like a champ. Got buyoff from CEO/CTO/CFO, deployed it; transitioned everyone over. One of the CEO's pet employees was pissed because of some problem with Mozilla/Firefox not working the way he liked. So the entire system was switched back, the very same day. Mail was messed up for a few days as we didn't control our DNS, and that was added to the blame game. Fun times.

No matter how much buy-in you get, people can be unreasonable about making any accommodation for change.

My mind thinks hierarchically in everything I do so I understand that Google's "Search, don't sort" is difficult to adopt to.

(Note, FWIW, though I think we're on the same page - the "business processes" I'm talking about in relation to ERP are less interface/program-mechanic based, and more along the lines of "Who is authorized/must approve" "How do we run accounting" "What is our procurement workflow" "How do you calculate taxes" etc)

I used to put a fair bit of effort into curating and filing emails, documents, etc. Over the past decade or so, aside from making sure that I put things I wanted to hang onto over the longer term into Archive folders so they didn't get deleted, I mostly stopped filing things and figured I'd just search for them if necessary.

Doesn't always work especially if I don't remember quite what I'm looking for. But it's overall a reasonable tradeoff compared to filing a bunch of stuff, most of which I'll never look at again. It is a shift in mindset though.

Years ago, I developed a similar habit with email.

I like to get emails out of my inbox as soon as I don't need them anymore (so my inbox always only contains the email that still requires my attention). But I don't do any sort of serious categorization/tagging/etc. -- the cost/benefit to doing that far too high.

Instead, I have a folder for each entity that I exchange emails with, and move the emails into the appropriate folder when I'm done with them, with no further categorization.

"one-off" emails (marketing, registration, etc.) just get deleted.

" that implementing a new backoffice suite by a consulting company is one third or less technical / IT project; and two thirds or more business process / transformation project"

It's at least 50% sales.

A cynical comment that has some roots in truth, but really needs some metrics defined to be useful :-)

By time spent? Not on anything but the tiniest projects which are likely to be loss-leaders anyway.

By value added? For consulting company, sure, that's one way to look at it; for the customer, probably not :P

For the purposes of my post, pre-sales/sales/RFP/whatever were assumed to be done prior to start of implementation.

Though I will agree that how ERP project is pitched/sold may well have an impact on the delivery:

- the IT vs BT categorization per above is likely implicitly or explicitly setup at these early stages

- who the stakeholders/champions are / who signed off on purchase and why - the CIO because their previous product is EOL? The VP of functional department (HR/Finance/etc) because they want to realize value in changing processes? Etc

- are the timelines and deliverables reasonable

- is the goal / value realized / ROI well defined and agreed

- are the requirements and scope well defined and agreed

- etc

(edited to change nested brackets into bullet points :P )

I'm not being cynical. The lifeblood of major consulting is the sales life cycle, which includes branding, relationship building. Lobbying costs. Responding to RFPs. Advertising. Trade shows. Expensive offices (to give the appearance of credibility). White papers. Paying off Gartner etc. so they put you in the 'good quadrant'. Partners at some firms spend nights a week having people in the industry over for dinner, nice wine etc..

Having been on the buying side of large enterprise projects (up to $50 million) I was surprised at the number of vendors who simply look at a very small outline of the project and decide not to get involved.

Later on, back on the selling side, I now appreciate that "qualifying out" opportunities is definitely a critical activity otherwise you well spend huge amounts of money chasing stuff that you probably won't win or, worse, win and make a loss on.

This goes back to my old quip of the easiest way to increase GDP to 5% would be to force everyone to learn how to actually use Excel. HT "You Suck at Excel" - JS https://www.youtube.com/watch?v=0nbkaYsR94c

Agreed! But the trouble for me with Excel is that it has an incredibly low ceiling. There are so many people out there doing what is essentially app development but with terrible tools. Tools that give them the wrong habits and mental models for working at scale.

I would love to see the spreadsheet reinvented to be a) collaboration-oriented, and b) a good on-ramp to ever-growing programming skills. So that when Bob in the next department over makes that crucial internal spreadsheet, you can safely and usefully interact with it from your crucial internal spreadsheet. Basically, to make every spreadsheet a potential microservice.

I've spent some years failing to figure out how to make that work. So if anybody manages, please let me know.

I have thought about this myself too, so I'll offer my two cents: Spreadsheets are really versatile, but they're a victim of their own versatility. You need to build something that is similar enough that people understand the concepts, but different enough so that people don't get frustrated by the fact that it's not Excel/Google Docs.

I always envisioned something that functioned like a Jupyter notebook, and pulled some pages from the MatLab debugger tool ...kind of like how Light Table implemented them, but don't get too caught up in that analogy. That way, you could create and define tables that are stand alone objects, and run [code] operations on data in those objects to create new tables. Everything that is processed from the tables is it's own [new object] entity, and can be [output or not] viewed or hidden. It would move away from the idea of having few tabs of large continuous spreadsheets in favor of having many small tables [matrices] that would have identifiers [variable names]. And if the user really wanted, they could drag around tables on a spreadsheet like grid, but the data in those tables would not be editable in that view. Oh, and graphs are their own entity too. Users can use a GUI editor to make graphs, but the graph properties will be user accessible in some sort of markup language for the edge case where the user needs to make a ton of graphs, and then needs to make some styling change to every graph. Not something I've ever experienced though... \s

This will:

1. Reduce the need for Excel's reference gymnastics that are performed when you move or insert rows/columns in to referenced data.

2. reduce the headache that is excel formulas. Since each object will be a formula instead of every cell. Also, the formulas don't have to be a single line, so it will be more readable.

The trick is finding a good language to use - Julia maybe? and designing a good GUI for helping non programmers do useful things like accessing and referencing data from objects in an intuitive way. This could be where the traditional spreadsheet view could come in handy. I like how you mentioned that it needs to serve as a good on ramp for ever growing programming skills, because it needs to be intuitive enough that non programmers can understand the basics and accomplish something useful, but powerful and hackable enough that the mental models can transfer.

We're definitely thinking along the same lines. I agree entirely about many small spreadsheets and the like. Another inspiration for me here is the long-lost Lotus Improv, which instead of being a blank page was more like a table, with explicit, named rows and columns. For me a general principle is to look at what people mostly do with spreadsheets and make explicit support for that.

You're very right that choosing the right language is important. As much as I love the languages I know, most of them are incredibly fiddly. E.g., things like dealing with "if a=b" are fine for those of us steeped in the mysteries, but a nightmare for more casual users. I'll have to check Julia out. Thanks for mentioning it!

There was also something called "Framework" that went a bit farther than spreadsheets. Of course when I write Spreadsheet here I mean "Spreadsheet products available in the 80s": https://en.wikipedia.org/wiki/Framework_(office_suite)

If you haven't used it, Google Sheets is worth looking at closely. It's good for collaboration (within the boundaries of a "normal" spreadsheet interface), and has a lot of options for integrating with external systems

I have used it, and it's definitely a step in the right direction. But it lacks a lot for me, in that the meaning of the system somebody builds is implicit and contingent.

E.g., in an OO language, I can be very clear about what a MailingAddress is. I can say what makes it well-formed. I can give rules about how it can be changed, and what happens when it does change.

But in a spreadsheet, that concept is mostly implicit. A human can look at it and say, "Oh, the mailing address is columns C-G, starting at row 2 and down to just before the first blank line." Until somebody edits the spreadsheet and inserts a phone number column, anyhow.

That means a Google Sheet can't really serve as a reliable microservice that knows about all the customer addresses, letting one's colleague's sheets read and update things. Sheet authors have all the necessary domain knowledge, but not enough of the technical knowledge, so at some point they have to bring in a professional programmer to re-express the domain knowledge in a language inaccessible to the people who formerly had control over their data and business processes.

For me, Google Sheets don't have the signature features of Excel, it's like Excel lite.

No VBA, no keyboard shortcuts, (no cross-workbook references).

Imho two people working inside the same worksheet at the same time was never an Excel's missing feature.

> No VBA

FWIW Google Sheets does have the ability to write macros in JavaScript. They even have an npm tool [0] that will let you edit offline via your preferred IDE and then push to your sheet.

[0] https://developers.google.com/apps-script/guides/clasp

I'm not saying JS is better or VBA, but JS is definitely not VBA.

Gsuite uses 'Apps Script'[1], not pure JS, so it is quite similar in function to VBA.


Have you tried AirTable?

I can imagine though, that there's a real chance that the "fear" of people having their jobs automated would cause them to cut back on spending, perhaps causing an immediate recession!

Perhaps it should become a part of public middle school curriculum?

... along with how to write a good search query

Funny enough, I had a class on this in grade school, taught by the school librarian. It included quoting, AND, NOT and OR style stuff you could use in library search systems and it is pretty much all irrelevant now, but it was cool at the time and I wonder if it helped to be exposed to that kind of boolean logic early.

Same, this was common in the late 70s and early 80s. Sure, you were on a mainframe terminal or a PC with a cd-rom, but the search principles are the same.

I was thinking more about the thought process for how to come up with the right words to begin with and for refinements to make to yield better results rather than the boolean operators although I think they are good for the young minds, too.

Funny you should mention this; we learned this in high school, 2 decades ago now. Quotes and OR and all. The task was to find some info on the teacher. It was in "computer science" class. Good fun.

I got an F.

Excel class was actually party of my junior high/high school (Netherlands). As well some programming in VB and PHP.

>Global 2000 company IT needs.

Splunk is a good example of this. I've worked at several public and private muti-billion dollar orgs, and they all rant and rave about how good Splunk is over whatever else they were using.

So, build more splunks. Maybe something that aggregates monitoring multiple servers. A plug and play Status Dashboard for a company's internal apps. Like this: https://www.google.com/appsstatus#hl=en&v=status

Splunk is great because they managed to anchor it to security as a SIEM and have an inscrutable billing model.

Which potential security threat do you want to eliminate to reduce the splunk bill?

Splunk is good, but it is expensive too, and given how - as you wrote - widespread is seems to be in big corporations, it must be a gold mine.

If you knew the technologies/software that Fortune 2000 companies are spending the most on, you should build complementary products, or similar things that solve the same problems those products are solving?

I think there's value in both. Consider Duo. They took a common problem - authenticating servers and services with 2FA - and made it easy to add to existing infrastructure. There's not really a consumer use for this product, aside from maybe a VPN, but there's absolutely a huge demand for it in business.

I worked on many CRUD apps that had say 4 different groups of users. We made a library / system for determining what "level" a user is, based on what Active Directory roles the user is assigned. In the application, you can show or hide components based on this level. Making some sort of middle-ware that abstracts things for developers, so they don't have to think about things like AD, is a huge area of opportunity.

There's no way to determine the total spend for specific software/technologies for Fortune 2000 companies (unless you can somehow survey them, or have an insider).

So what proxy method would you use to determine that Fortune 2000 companies are spending twice as much on Splunk as they do on New Relic (for instance)?

Duo is awesome. I discovered it because it’s the system my employer uses for 2FA and push notifications are so much nicer than having to enter codes.

But similarly, I’ve found it convenient for all my other 2FA needs from GitHub to RuneScape. So even then, it has value to consumers.

This is the growth angle for pretty much all of the IDAAS (Identity as a Service) players in the market that are not AD. Will be interesting to see how it evolves.

The tricky part of that is the loading and mapping from the companies representation of the data to the SaaS representation of that data to run whatever analysis the product wants to deliver. That step can be almost as much work as just writing the application from scratch since now you may have 2 copies of the data you need to keep in sync, the actual copy the company uses and the copy used for the SaaS analysis (and often multiple different copies for each SaaS the company is trying to use to substitute for just writing the software themselves).

I worked briefly with a startup who's idea was to create an index of your entire corporate data (PDFs, powerpoints, docs, etc) and their product would intelligently serve up documents as you typed.

So for example, if you would pen an email/slack talking about a slide deck from a recent meeting, in the widget the product would have already found the deck ready for you to drag and drop it into the message.

It worked freakishly good, and saved a whole bunch of time trying to find a file in the mess that everyones Google Drive becomes.

The problem they were having was no enterprise wanted to hand over an entire copy of their data for them to index.

AWS just introduced Kendra to do the same thing. They will likely suffer the same concerns. Google had a search appliance that I thought was a brilliant idea but never got much traction: https://en.wikipedia.org/wiki/Google_Search_Appliance

Google Search Appliance was used by thousands and thousands of big corporations - so to say it didn't get much traction is not correct. For many years it was the defacto solution for enterprise search solutions.

Huh, I basically never really heard of it being used. Guess that's the trick with 'enterprise solutions' outside of explicitly looking for it or it popping up on an aggregator like HN or /. (since this was starting in the mid '00s) you don't really hear about them.

That's another difficulty. Even if the actual transfer is relatively painless it is a risk every time your company has given a copy of customer data to another company.

Using AWS for example all it takes is a misconfigured bucket to expose massive amounts of customer data.

At least it is private by default now...but the point stands.

We basically did that for healthcare (electronic medical records) in the mid 2000s. Our customers (actual users) loved us.

Like your enterprises, no hospital (or equiv) would consider exfiltrating their data.

So our gear was physically deployed and operated on each hospital's own system. Our marketing and sales goons called this our "federated data model", which had no tangible connection to reality.

I likened the experience to building model ships in a bottle, while wearing mittens, while someone is punching you in the face. To mitigate, we created what would now be called CI/CD pipelines. We managed 100s of deployments with near zero downtime and effortless rollbacks. I'm rueful just thinking about it.

This strategy didn't survive the 2008 economic crash.

There are multiple, mature products in the "enterprise document management" space that work similarly to the product that you describe, available on premises and in the cloud[1]. There's also (smaller, specialised) consulting companies that show how to implement these products into a business. There's some really nice features like robust OCR, meaning after scanning, you can search immediately for document metadata, like invoices numbers- for companies that insist on having paper documents as part of their processes, or crusty old management that refuses to have digital signatures.

The challenge is the same as the main topic of this thread- change management. You have to force employees to manage their documents exclusively within this system rather than dumping everything into shared drives and using Outlook attachments for everything.

[1] One example: https://www.alfresco.com/ecm-software/document-management

Google still offers an enterprise intranet search product which will index anything you let it access.


Was there no self-hosted option?

> The problem they were having was no enterprise wanted to hand over an entire copy of their data for them to index.

That sounds like a reasonable response. At least, I know that would be a showstopper for me, personally.

Microsoft teams is well-positioned to do this as it is integrated with SharePoint and OneDrive. However, it's not super easy to add metadata...

FWIW, robotic process automation (RPC) is very hot these days, and it's always a bespoke service - unless two companies happen to use the exact same software, and follow the exact same business logic.

The RPC consultants, often titled AI / ML consultants, make money hand over fist traveling from site to site and implementing some automatic procedures.

Same with chat bots. The past 2-3 years there's been an explosion in demand for chat bots, and these often take entire teams to implement and train. Again, lots of bespoke products.

Who gets these gigs largely depends on networking and sales. The big companies (Accenture, etc.) are rolling in this, since they also have a good picture of company software from previous projects.

The accounting industry is also desperate to get in on the ML revolution. I've lost count on how many times I've been contacted by recruiters in that space, who want magic ML pixie dust.

The market is there, but it can be difficult to break into as an outsider, or smaller player - if you're planning in building some one-size fits all product. Best plan, IMO, would be to focus on some niche part of the process, and become the best at just that - and then pitch to the big companies that deal with actual F500 clients.

There are of course exceptions, but a 5 man startup vs Accenture offering the same base product, Accenture is gonna walk away with the gig 99 out of 100 times.

(But with that said, a lot of the startups that focus on these spaces, seem to be ex-consultants from said big consulting firms. Logical enough, as they have both seen the problems first hand, and built networks within the industry.)

Just a small correction (sorry if I sound pedantic), I think you meant RPA instead of RPC:


I noticed that since I worked at the RPA team of my current employer. :-)

> I would love to see the following analysis: A map of repetitious tasks, spreadsheets, and manual data extraction by function in the Fortune 500.

This is basically asking for a market gap in a potentially very profitable market. That's silly, everybody wants that.

ERP is hard. VERY hard.

It is, but there is also a lot of artificial complexity created for lock-in purposes.

A lot of ERP systems move away from well-trodden programming tools to their own custom stuff. I know of ERP systems have their own source control version control systems. Why? They're often awful to use. The programmers that use them are often business-focused haven't really used stuff outside that ecosystem. So they don't know how bad it is.

Most ERP implementations i've come across make little use of automated testing as well.

or... they create their own internal DSLs - customers and internal people use those - and everyone spends years thinking about the 'custom DSL' way. Developing any sort of automated testing around your own custom language basically becomes impossible.

I did some work for a place where, they embedded groovy as their 'write custom rules stuff from the UI' language - that was at least easier to reason about, and to debug/test offline (not automated, but somewhat easier than a DSL that only runs inside the system).

Automated testing is becoming the norm, but I'd agree that otherwise ERP developers, on average, are either unaware, or simply not able to utilize most of the tools and practices from the wider market.

I believe ERP SaaS would fall in the category "falsehoods programmers believe about how companies store their data".

There is always this exception that makes a SaaS ERP not fit for a company.

I've gone down this rabbit hole, it's amazing how often orgs reinvent the wheel. I don't fault them, there's not support and guidance on this sort of stuff & incentives aren't aligned for product & development teams to focus on their corp's primary value prop vs basic stuff like user management and customer service tooling.

Hoping to get a white paper or at least a one-two pager in the next few weeks.

Shameless plug, follow here for updates if this sort of stuff interests you: https://hipspec.com/

Sometimes people reinvent the wheel simply because they need a new project to take credit for. Every link in the chain needs an accomplishment to point toward when justifying themselves to superiors.

"Budget breakdown of current software spend, by function, by line, in the Fortune 500."

Just curious if the OP wants this so he can build a product FROM this data, or if he just wants this data as a product, alone.

How about turning those into open source projects? Especially across public organizations who could develop the systems together if they just knew they shared the same needs

How would this be a product? Isn't this just research?

Why would a Fortune 500 give up such information, which is pretty much how they make money.

Also, if they did map out any tasks that could be automated, they would automate them themselves.

Good idea, but not a product.

Let me know if you find it

#2. I've had an idea I've mulled around for a couple of years now about a new social network. One which limits your 'feed' to 25 friends and that is it.

25 might even be too high (or potentially too low; the number is arbitrary), but the general idea is that in your life there are only a handful of people who you should really care to keep tabs on. You can "friend" more than 25 people but you can only see the activity of 25 of your friends and the others are relegated to essentially being contacts.

I remember about two years ago Facebook was getting into a lot of trouble due to the amount of negative mental impact it had/has on its users and they did a study which found social media can have a negative impact on its users unless the user only follows a small circle of real-life friends and interacts with those people online. I don't have a source to this; I'm willing to accept there isn't real _evidence_ behind this claim.

Anyway, I'm surprised social media has been around this long, has been the subject of such controversy and there haven't been any or many real, impactful changes to the nature of it. Facebook and Instagram are, in my opinion, trying to be too much. They want to combine personal and public spaces. This, of course, works from a business perspective. But what are the long-term consequences of these products on our mental health and society?

We need a new social media focused on personal, tight-knit groups of people and interests. One that makes this the focus and doesn't stray for the purposes of profits. And, if that's not feasible in an economy that demands growth, we need better legislation demanding certain consumer protections are created for this sphere of products.

I can already see it, imagine people who have their list full, removing people and replacing them would be a public passive aggressive "thing" people would do and the such drama that would follow. I'm not saying this is a bad thing, I think this would actually be a selling point of the platform, however could possibly turn into a negative part of the social culture.

So Myspace top 8 list? That's exactly what it was. Weade a game out of adding and removing friends from our top 8.

Some of those friends didn't realize it was a game.

What if it automatically populated the list of your "close friends" by some algorithm that ranks your overall interaction with them. So you can't arbitrarily add or remove the friends. You could "game" it but only by actually interacting with the person. If they include sentiment analysis on the interactions, then you couldn't "game" it by spamming or being a cyber bully either. You'd actually have to be friendly towards that person.

If I want to person A as a close friend, and they're not active on the platform, that should be allowed. Maximizing interactions isn't the goal, or else we wouldn't limit the feed in the first place

you just described facebook's feed algorithm

I think it can built so the "Feed" is restricted to "Close Friends". You can still go checkout the other friend's profiles individually. I'd also like it to be set to filter out news articles and other crap like what photos they liked ...etc.

One which limits your 'feed' to 25 friends and that is it.

An ex-Googler (that never played defense![1]) tried that with the social network Path. I think the limit was 50 friends though. It failed.

Edit: [1] Quote attribution: "I don’t use a ring of any kind on my phone. This is so that I am always on offense and never defense." - Dave Morin

To be fair, almost every non-Facebook social network has failed. I wouldn’t consider that to be proof that a limited social network can’t succeed but rather that social networking sites are difficult to produce.

No disagreement here. I just recall that guy being comically pompous (see also his penchant for carrying both a "day iphone" and a "night iphone") at the time Path launched.

Wow, what is it with anti-social people starting social networks? I long ago lost track of my collection of ridiculous quotes from the Friendster guy. But I still remember him being outraged that people were using it wrong by creating profiles for abstract things they loved (cities, parks, stores, brands) and then friending them. Like, buddy, when your users find new ways to use your product, run with it. Instead, he just got big mad and banned a lot of people.

Looking back, Friendster strikes me as the single biggest missed opportunity of that decade. He had a two-year lead on Facebook. In a network-effect business! But through careful focus and diligent effort, he managed to blow it.

To some anti-social people, building a social network might seem like a solution to make social interactions more manageable.

wasnt inability to scale a part of it tho? we didnt know how to architect for eagerly connected graphs then.

That's why this has remained an idea for me rather than anything I would consider attempting. Social networks have an enormous bar to entry. This is why I think this would most feasibly be applied as a restriction to Facebook/Instagram and it's probably something they will never implement unless forced.

Facebook and Instagram have the concept of “close friends”. A lot of people use it pretty close to how you’re envisioning.

Idk, I wanted to try path and never even got the chance because it never launched as far as I know (for the web at least?)

Have we ever seen social networks succeed that were not primarily driven by highschool kids? (except work oriented ones, e.g. LinkedIn/tech communities and pre-www communities)

> We need a new social media focused on personal, tight-knit groups of people and interests

Group chat: text/Messages/WhatsApp. The fewer frills, the better.

IRC is still kicking in some communities!

I've kind of achieved this for my Instagram feed. I simply "mute" anyone I don't care to see again. My account still follows them, so they are none the wiser. I think most social networks provide a similar "publicly follow but mute" feature.

It takes some time to mute everyone I want muted, but it only took a couple days to mute the more post-happy folks. I discovered that most of my network hardly post.

Instagram is so much more enjoyable now. 2 minutes a day, rarely more.

>there are only a handful of people who you should really care to keep tabs on

I don't consider myself an extreme socialite, but as someone who has lived in 3 states in the US and whose friends are scattered around the world after getting a degree here, this sounds really, really, really perverse.

That said, most of my use of Facebook is not to keep tabs on the people I care about (a lot of them don't use Facebook much, if at all).

The product that the OP described already exists: it's called a chat room (IRC channel, Skype group chat, Telegram channel, Whatsapp group chat, etc).

Reminds me of Google+ circles

I really liked the idea of Google+ circles. Creating "containerized" groups of people (things?) that I could act on in a complete unique way was really cool. I could say something in my "friends" circle that I couldn't (or would be misunderstood) in my "family" circle. I could also keep track of technology (Linux, Tesla, etc.) in another, completely separate, area.

I wish that a "social" network would bring back the idea of Circles.

Dunbar's number seems relevant for this line of thinking: https://en.wikipedia.org/wiki/Dunbar%27s_number

Thanks for this. I'd never heard of it.

Private/restricted social networks do not work as businesses. They are antagonistic to both viral growth and hit content.

Many, many have tried.

This is exactly what I'm aiming to do with Thread - an alternative social network that focuses on tight-knit groups and building more intimate relationships, for the purpose of improving our mental health.

One important thing to note is that it will be monetized through a subscription, not by selling consumer data or driving sales of targeted ads.

Check it out @ https://get.thread-app.com/

Can’t that be achieved by simply Facebook offering an option for feeds of manually created group by you? Like you create a group called closefriends where you add your close friends and family upto 25 members. Then a feed for those people only shows up?

Almost similar to where each user is a subreddit and you create a multireddit.

Facebook is a video game where you honor the groupthink for max followers and upvotes; something like the 25 followers can be emulated but if you're the only person playing the new game, its not going to be a fun game because everyone else will have higher scores due to no made up limitations.

I do this right now on Facebook; you can unfollow anyone. I'm not following anyone and I have no feed. But the problem is people aren't going to voluntarily do this, and it's a big pain to unfollow hundreds of people.

This needs to be baked-in to have any real impact.

Isn’t that the premise of twitter?

That reminds me of Google Buzz. I used it with a group of maybe 5-6 other people, and it felt like Google had built this product just for us- deep integration into Google Chat and Gmail, everything was solid, and no one else was using it, so it felt like our own private ghost-town.

> I've had an idea I've mulled around for a couple of years now about a new social network. One which limits your 'feed' to 25 friends and that is it.

C.f. https://cocoon.com/

I think this gets pretty close to what you're after: https://runyourown.social/

I don’t know about Facebook, but in its Russian analogue it is possible to remove your friends (and groups) from your feed.

This is exactly what Path tried to do.

You can easily do this today, just start a mastodon instance and restrict the signups. I run a > 100 user instance on just a $10 digital ocean VPS. I'm seeing small tight-knit community instances popping up all over the place. You can tweak the settings or or even add customizations from forks that cater more to a siloed instance.

We need to separate the data layer from the app layer to allow free competition again.

I don't think that Facebook will be replaced with something so similar. I think what we need is an open data-platform that can contain (in a controlled way) social media data along with other kinds of data and then allow building social media apps on top of it. Otherwise we'll always be stuck using whatever the biggest supplier builds for us.

Nah, this is what Facebook built and it caused Cambridge analytica.

"controlled" is a key word here


>What would be a network which allowed for more thoughtful discourse? Or at least the ability to more actively mute topics, threads, and groups of users while surfacing better content algorithmically?

I would be interested in a parliament-like protocol for discussion, structured debate and reaching consensus.

We currently have tools for proposing changes to text documents (such as pull requests) which could be applied to a community rules/laws or values system, but we don't have anything that can debate the changes, log objections, track resolutions and compromises in a structured way.

I've seen places where community decision making is emergent (e.g. autobans when a users reputation drops below a certain threshold or votekick in games) but nothing that formalises the process, allows review of results, links decisions to overarching principles (or notes where a principle has not been followed due to circumstances).

To add to this concept, it would be great if the elements of a quality argument[1] (i.g. Claims, Counterclaims, Reasons, Evidence, etc.) were built in. Also people could flag a comment (or parts of it) with a specific logical fallacy[2].

It would help guide people into arguing more fruitfully and better digesting what they read.

1. https://study.com/academy/lesson/parts-of-an-argument-claims... 2. https://yourlogicalfallacyis.com/

It would be great to show arguments as processes (data, model(s), evaluation/claims), and other arguments as specific problems with the process (logical problem, bias, low quality data, bad statistics, etc..)

Ultimately everyone should be able to agree on the data, the model and the evaluation separately. The differences should boil down to differences in personal assumptions (different Bayesian apriori distributions, eg. different morals, worldview, thresholds for things, etc.)

This should lead to questions about stuff that is either subjective and thus people can agree to disagree, or to simpler objective problems. (For example that data collection methodology leads to higher uncertainty than claimed, hence the whole argument becomes inconclusive.)

It's not always possible to reach consensus. Sometimes you just have to admit no one really knows and leave the problem to experimentation or further inquiry. Or maybe it's just undecidable. But most people don't have the humility you admit that.

What's missing is organizing and breaking down big questions in a structural way. And naturally quality controlling smaller parts should be easier. (Especially if QC itself becomes structured, eg. show someone a simple argument and ask whether it is biased or not; then meta check the bias moderators and update weights accordingly. This will probably show polarization, so maybe then that might allow to algorithmically pick sane centrist moderators.)

Won’t votekicks lead to echo chambers?

Every single system you come up with to stop trolls and other abusive behavior will enable echo chambers. Even upvote systems.

> NoCode/LowCode (letting anyone build an app with a spreadsheet as a database representation)

AirTable has made huge inroads here. I'm also building something in this area (https://lightsheets.app/), turning slightly back towards spreadsheets instead of databases but then building enhancements and modern integrations on top. I think there's still lots of potential in this area, despite spreadsheets being 50 years old.

Airtable did some great work on the UX. Introducing people to table schemas with typed columns is pretty awesome, and the creative way they've leveraged that with the little tab that comes out of the side for either a nice view over the data (e.g. a map for an address column) or for showing Pivot Tables/Aggregations. They deserve some design awards for those ideas.

But it's not a great alternative to a DB. It's really expensive per seat (twice the cost of Google Apps and you only get one app). The row limits on tables are too low. And I'd really like native webhook support to ingest row updates.

I'd be very interested an Airtable clone that I can run locally on-top of Postgres or Mongo or something like that. My main use-case would be to replace expensive-to-build internal CRUD apps that really should just be spreadsheets but require bespoke integrations with other internal systems.

Airtable has done a great job with their UX!

I've started working on a similar platform where the goal is to make building CRUD apps wildly simple. The plan going forward will be to layer on more sophistication after the core foundation is extremely solid.

The platform uses Mongo for the dynamic data tables which is a game changer for tools like this. Postgres jsonb data would also work, but Mongo just feels more natural to me.

I've named the platform Webase. Check it out here: https://www.webase.com

Note: this is very basic currently but consistently getting better!

Hey I made an account but only see

> You have not created any apps yet.

with no way to create an app that I can see

Retool seems like what you want!


Awesome. Thanks for Validating. I'm going to spend the next year of my life building exactly that. Right now I've just put it a manifesto with some ideas at http://orows.com, but will be coming out with a proof of concept very soon.

Cool! I'll add one more thought that I think might be particularly useful

> Granular permissions not only at row level but field level too. You can share parts of data with others for editing or have it view-only with ability for commenting and accepting suggestions

In my org (and I suspect many), many of the Sales/Operations people maintain Google Sheets with links into our internal apps. Generally it's TODO lists or other little organizational/workflow type stuff

What I think makes Airtable so powerful is that foreign-keys are first class. You can create a table and include references to rows from other tables. This allows the end user to create tables to support their workflows directly in the same tool they use to manage the data, and means I don't have to build that same functionality into the CRUD app (and until you're deep in a workflow it's hard to get it perfect, so there's always back-and-forth).

A good permission system would allow me to create a set of "core tables" that I tightly control the schema of, but allow others to still create tables which reference this one.

I'm interested in link density and diversity in those little workflow supporting spreadsheets you mention.

Do you have some examples?

We're building a sort of highly linkable (cell level) spreadsheet-type widget builder that is supposed to play nicely with existing workflows (instead of trying to force people to build out entire workflows like many of these new breed of "CRUD-replacer" low-code/no-code apps seem to want us to do). Might interest you!

I can't share anything directly but let me describe a little scenario for you that's close enough:

We have a CRUD app that acts as a CMS with all our product listings in it. There is a team of people who are tasked with curating and maintaining these product listings. So you can go to crud.com/admin/product/<id> and modify the listing, which needs to happen every now and then.

The team of people who maintain these product listings need to coordinate their efforts and so when they need to modify some set of the products they create a spreadsheet with (product_id, link_to_crud.com, Person who is responsible, status). Maybe there's some approval process or multiple statuses or something like that that makes this all a little more custom/complex. Maybe they need to update a few things across services at the same-ish time so there might be a few links per row but it's going to be relatively bounded by what a human can handle.

If our data lived in Airtable, it'd be really easy for this team to make their tables/spreadsheets with direct relations to our core tables (which replace the CRUD app in this case). That team understands their workflows better than I ever will as an engineer. Giving them the power to create their own workflow tools is way better than trying to have engineering build it for them, and that's going to be a huge selling point for me.

I don't think we'd need cell-level links but rather row-level. Right now they use Google sheets, but a more domain-specific tool would be better.

That makes total sense. Yeah you want to see tables with relationships/hyperlink to other tables. Deffo don’t want to be manually joining ids everytime.

It's more than that though, it's that I need to be able to create "core tables" that only privileged users can modify the schema of, but then let anyone create tables with relationships to those core tables.

Glide apps also falls into this realm. You make an app from a Google Sheet. Found out about it from HN



I think building apps with spreadsheets makes sense, but to me the spreadsheet works better as the UI, not the database. In other words, you still store your data in a plain RDBMS, but you access and manipulate that data via spreadsheet formulas or custom functions.

Maybe... but a regular spreadsheet doesn't map very well to a RDBMS due to its variable and flat structure. It might map to a column data store like Cassandra I guess, wonder if anyone has looked at that?

What I'm working on though is more about taking the core of existing spreadsheets and building powerful integrations on top of it. Like in the sibling comments, building a webapp, or something else. I really think this has huge potential.

What about a spreadsheet program that can be scripted with Javascript?

JS is nice because all you need to write it is a text editor and a web browser, both of which just about all computers have. I had an entry-level job that involved a computer once and automated a lot of the painful stuff with JS - I bet that if more people knew JS, more people would do the same.

I don't know if that'd be a good business proposition, but I think it could be a good on-ramp.

Google Sheets uses JavaScript as its scripting language. The problems I've had there are related to how complicated it is to get data in and out of Sheets. If you want to update a sheet you own, you need to do a full blown OAuth2 implementation.

Excel has an JavaScript API now as well right? At least I heard rumblings to that effect. Not sure how hard / easy it is to get data into / out of though.

That's what we are building over at https://clay.run (free preview at the moment). We built a programmable spreadsheet.

Been using Clay for some while. It has some nice advantages over Airtable. The developer experience is much nicer with unrestricted row colums, full sql query etc. Does a better job at bridging the gap between technical requirements and business requirements.

Every other week, someone releases a version of spreadsheet-as-database on Product Hunt. This is one of those “sounds great for engineers” memes, that’s actually not that high demand IRL

source: I made one myself (and know the founders of others)

AwesomeTable too

#4 is a no-go in my book.

Things like Ring, Nextdoor, and Facebook already exist. Nextdoor and Facebook are riddled with inanity, from political rants to dumb jokes to hoaxes to common scams to law enforcement rants.

I don't want more information from my neighbors -- 99% of it is garbage. I want highly-filtered information. Basically, a neighborhood watch but with an aggressive spam filter. Right now I glance at the various neighborhood feeds with one eye closed, sifting through the intense stupidity, trying to capture valuable intelligence.

Of course, all of this may be a dumb corporate-run idea and maybe people should really focus on forming good relationships with their neighbors in meatspace and talk to them in person aka HUMINT.

The "SaaS-ification" of security cameras should also worry us. Ring — not its customers — controls the videos taken with their cameras, creating a video surveillance network that the police can access without needing a warrant. They've partnered with 400 police forces to give them access to that data [1]. Although they claim to let customers deny police requests for footage, their terms of service allow them to hand video over to police if they deem the request "reasonable".

And it's not just Ring customers that are affected, it's anyone in the general vicinity. If your house is in the field of view of a neighbor's Ring camera, you're being surveilled too.

[1] https://www.washingtonpost.com/technology/2019/08/28/doorbel...

Things like Ring are really worrying because it's Amazon. Fear is the motivator in security camera sales. With Ring you've got a parent company whose CEO owns a national news organization(Washington Post) and a large media company(Amazon). Amazon is so aggressively pursuing being in every part of your life that the fear motivator is going to be really easy to manipulate.

Here's is a great long-form article on some of the societal implications of living in the age of Ring and NextDoor. It touches on this point throughout the article: https://www.theatlantic.com/technology/archive/2019/11/steal...

> Although they claim to let customers deny police requests for footage, their terms of service allow them to hand video over to police if they deem the request "reasonable".

Defining what is a "reasonable government request" is a valid question, but it's really just not that high of a bar to get a records subpoena/search warrant for video like this. Courts sign off on those routinely, so I don't think you can really expect Ring or any company that holds your records to deny police requests for very long.

The system they have seems pretty balanced. The police look at the ring website to see who has cameras (they could figure that out by walking the neighborhood), they ask for the footage (instead of knocking on the door), they get turned down (or not), they get a warrant, the footage is released. Ring is reducing the overhead of asking somewhat, but they're not enabling mass surveillance or building AI systems that track suspicious people across multiple ring devices.

Is there something I'm missing here? If you record video of your front yard, and the police want to see it, they have a right to, subject to the normal judicial review.

> Is there something I'm missing here?

Yes. If Ring had even the barest shred of ethics, they would client-side encrypt the videos stored in their cloud. It would use no extra space. The user would have to explicitly approve the decryption and sharing of videos.

But they don't, because being able to access and datamine those videos is a huge money-maker for them.

Ring is terrible, but their engineers willing to implement this corporate surveillance state? They're the worst. They wield their software skills as a mercenary would a weapon against innocents. I seriously cannot even comprehend how they sleep at night and look at themselves in the mirror in the morning.

At least folks over at Apple are sane and are making sure that HomeKit surveillance videos are client-side encrypted.

And then the second someone lost their phone, their camera becomes worthless and they can't view their videos.

The average user can't be trusted to manage client-side encryption keys reliably.

No? You can derive a key from a password.

> Is there something I'm missing here? If you record video of your front yard, and the police want to see it, they have a right to, subject to the normal judicial review.

The issue is that Ring's terms specifically allow them to circumvent "normal judicial review" if the request is "reasonable". From the same WaPo article:

> Ring users consent to the company giving recorded video to “law enforcement authorities, government officials and/or third parties” if the company believes it's necessary to comply with “legal process or reasonable government request,” its terms of service state.

I'm fine with the police having access to video after obtaining a warrant or subpoena, even if it's not a particularly high bar to clear. But that should still be the bar. We shouldn't expect Ring to refuse police requests even after being served, but we should expect them to hold out until that point — and unfortunately, we can't trust them to do that.

> Ring is reducing the overhead of asking somewhat, but they're not enabling mass surveillance or building AI systems that track suspicious people across multiple ring devices.

Are you sure?

How can we as citizens verify it?

> Are you sure?

I'm not. Do I think it's likely that they are? No.

> How can we as citizens verify it?

The same way we verify that Google isn't producing broad-scale AI systems looking for specific subsets of people across the GMail data. Investigative reporting, whistle-blowers, regulation/lawmaking, and looking closely at the evidence presented when the government acts. This is why parallel construction is pernicious, as it prevents meaningful oversight of government malfeasance.

End-to-end encryption, and user ownership/encryption of data is also great, but it's not widely available, and many use cases don't work when the service provider can't see the data they're storing. Even when the data is encrypted, you can get a lot of valuable intelligence from metadata.

So, in other words, citizens can't verify anything. Instead, all we can do is hope that any abuse will eventually be noticed and reported by some random whistleblower somewhere.

That's hardly sufficient, and especially not with a company like Amazon.

Although I find your answer upsetting, I also find it reasonable. +1.

> Ring is reducing the overhead of asking somewhat, but they're not enabling mass surveillance or building AI systems that track suspicious people across multiple ring devices.

It really looks to me like this is exactly what they're building.

My ring is basically useless for surveillance, at night especially. It can't see across the street and barely gets beyond my front stoop. This is even with the lights on and a street light nearby. Likewise for my Arlos, useless at night. During the day they are a bit better but the resolution isn't anything special and it'd be impossible to catch a license plate. Make, model and color but that's about it.

In the UK, you can't lawfully film another persons dwelling, in that you would break the law by having any of a neighbours property in view.

No, this is wrong. There is no law that explicitly prohibits filming the public realm (e.g. the pavement) or incidentally capturing your neighbour's property.

There is a code of practice: the Surveillance Camera Code of Practice (SCCoP). There are also requirements to follow under the General Data Protection Regulation (GDPR) and the Data Protection Act 2018 (DPA).

The consequences of not following the SCCoP, the GDPR, and the DPA may result in regulatory action being taken against you by the Information Commissioner's Office as well as private legal action by the affected individuals, but it is not per se against the law to film those areas.

Yeah, that sounds horrible... if it is anything like Nextdoor, there will be three posts a day about "There was a black guy sitting in his car!"

Nah, you are thinking way to short-term.

The AI will flag the black person himself as often associated with complaints and posts tagged "danger". Thus barring him from job interviews and loans by other automated services (Uber, but for HR!) who bought the data set "to elevate user experience".

Creeping right back to apartheid, but I would say entirely worth it if it means I can get a notification that my cappuccino creamer was stolen 5 minutes ago by an unidentifiable person! /s

He was sitting there...menacingly!

I cancelled my Nextdoor account in part because I got tired of all the paranoia and inevitable devolvement into political bickering, but also in part as a protest against Nextdoor that they need to police this kind of activity. It was only afterwards that I realized that Nextdoor wants this activity. Like Facebook, they see it as encouraging engagement thus traffic they can use to sell ads.

It confirms to me that cancelling was the right move for me but it also saddens me that we've gotten ourselves into a state in technology where the business model is to get people riled up and then profit from that discord.

NextDoor is great because I can learn who is and isn't racist in my South Carolina suburban neighborhood.

> Of course, all of this may be a dumb corporate-run idea and maybe people should really focus on forming good relationships with their neighbors in meatspace and talk to them in person aka HUMINT.

But then how do we filter out the political rants, insanity, dumb jokes, hoaxes, and common scams?

Thankfully, people usually do a decent job of filtering themselves in meatspace. At least, they are better at it in person than they are online.

There are also social cues and the lack of an ever present mob. A lot of people who are political activists online are not actually that confident in their views and absent the audience and ready defenders, won't express their views because they know they can't back it up in a one-on-one conversation.

Our city did a traffic survey of a local street that is the primary route for everyone going from surrounding neighborhoods to downtown and found the average speed was ~2 mph over the posted limit, with the 85th percentile at ~5 mph over the speed limit. These are reasonable speed for such a wide road. The city then added several 'safety' features to it, and found in a subsequent survey that the changes actually did little to slow traffic. One feature is a pair of speed bumps. Some drivers slow to about 5 mph to go over the bumps, increasing the risk of rear end accidents. Other drivers veer into the bike lanes to go partially around them without slowing down. Everyone else slows down a reasonable amount, but then subsequently accelerate back up to the speed limit on a block where jaywalking is common. Some of the features they added are, in my opinion, beneficial, like turning a side street into a one way road after frequent accidents from people turning left out of it.

But you can't talk about it on Nextdoor. The second you criticize the speed bumps you'll be ignored (if you're lucky) or attacked. Nuance is not allowed. You're either for all the changes, even the dangerous ones, or you're a crazy speed demon. Meanwhile, if you post about 'the children' who you are afraid for (even though no one has any evidence of a child ever being hurt by a driver on this road), you're going to get upvoted a lot within the first ten minutes.

Heh, a road to my parents’ used to be riddled with potholes and uneven sewer grates (Canadian winters + drainage issues probably).

The city spent months ripping it up and paving it nice and smooth.

Then a couple months later they put up speed bumps.

We spent a lot of money to get back to the speeds people already drove.

People on the next street over (it's a grid) are on Nextdoor complaining about all the traffic coming down their street now. And I'm sure it's the worst of the traffic that's going over there to avoid the features. I'm not sure how the city will fob off spending money on that street after spending the money on this one.

I once lived on a street that was incredibly badly maintained. The city informed the residents that they would be repaving the street and the residents campaigned hard for the city not to do that, for precisely the reasons you're talking about here.

Their campaign was successful, in no small measure because a poll of the residents showed 95% of them opposed the repaving, and that the local paper wrote a big story about the whole affair.

I hate how pervasive this degenerate thinking is in local politics. "we don't want a nice thing because then people would use it."

That's not the thinking at all, though. The thinking is that the road being in rough shape means that everyone is forced to drive at a sane speed through a residential street where children are frequently playing.

In other words, the "nice thing" isn't actually so nice in terms of the things that the people on that street really care about.

> focus on forming good relationships with their neighbors in meatspace

Remember that garbage you mentioned. We don't want to deal with other people's garbage in meatspace any more than we do online.

Many people self-censor to a much greater degree in person interactions vs what they post publicly online.

No man's an island. What some post online is nothing but free-wheeling, disconnected from feedback.

Everyone wants good editing and content as product feature. You can pay...

Products I'm glad do NOT exist:

4. ADT 2.0: Digital neighborhood watch.

Ring is gross enough, thank you very much.

My ex-neighborhood (Magnolia in Seattle) has had Flock Safety install 6 "AI" camera systems, two at each of the three entry/exit point of the neighborhood. Each installation has a camera overseeing entering traffic and one overseeing exiting traffic. The cameras have license-plate recognition and object detection for car color/make/model, pedestrians, bicyclists and even animals. Flock Safety charges $2,000 per year per camera. Magnolia is one of about 10 Seattle-area neighborhoods with Flock Safety cameras installed. The Magnolia residents themselves pay for the cameras and service. They all can log in to a web service and see the footage and ALPR lists, timestamps, etc.

Friends I have who live there are thrilled with the system although the Seattle police seem to have little interest in the system telling the residents that they are only interested in seeing footage that shows suspicious activities or is directly related to a crime.


Porch pirates are so bad in my new neighborhood that I admit to wanting one of these on my dead-end culdesac and have investigated what it would take to build such a system myself.

1. Set up some raspberry pis with cameras that run a webserver that when hit, it returns the current image.

2. Create a backend that queries each camera for their current image.

3. Run the images through some object recognition neural network (fastai has some great object recognition tutorials).

4. Make your backend stitch together the images from each camera to make a video of what was being recognized when objects of interest were detected (each camera will create its own video)

5. Create a nice little UI/App for everyone to view the movies with some filtering in place (time of day, object(s) of interest).

I want to create something like this as well. All of the pieces seem to be pretty straightforward to me except for figuring out power and network connectivity. I guess each neighbor would have to configure the device to use their SSID... and hope someone doesn't steal the raspberry pi and get the SSID credentials from memory... Maybe just make the neighbors use a separate network?

Wow. Thank you. There is an OpenALPR project for license plate recognition.


Also at: http://doc.openalpr.com/opensource.html

And a ready-to-go version you can buy/subscribe based on their stuff:


It already exists in apps for nosey neighbors like Nextdoor

I signed up to our Nextdoor (uk).

There seems to be four things on here: - an avon lady over agressively spamming her stuff (blocked) - people asking for small job recommendations (who should I get to fix my garden gate, babysitter, tv ariel fitting) - some notices about local events (when / where fireworks, Remembrence, carolling) - group commiseration that one time some group was speeding all over town at midnight running red lights and keeping us up with their skidding and revving.

Honestly pretty good. I see it maybe once a week.

I went to sign up for Nextdoor and it wanted full access to all my contacts in my phone. I looked on their site and that is so they can invite friends from my contacts. I didn't sign up, maybe i will later, leaves a bad taste in my mouth.


I have the iphone app and it didn't require this permission or maybe it asked and I denied it. Is that not possible in android?

I just checked my app, it had no permissions.

I wouldn't allow it to do so, so I assume it asked, I said no, and we moved on.

Aka: “I see a suspicious looking Black Guy in our suburban neighborhood. It looks like he is breaking into a house by using a remote to open a garage door and driving in.”

Yeah, my Next-door has a lot of that. They aren't that obvious though. I like to call them out.

"I saw a suspicious man walking down the street!"

"What made him suspicious?"

"He just doesn't look like he belongs here."


"He's black"

"Being black isn't suspicious"

I then I get an email from Next-door about my comments being reported!

Also: "why haven't the cops arrested all these homeless people yet?"

The only value Nextdoor provided me was making me aware that the asshole down the block is also a racist.

Deleted that cesspool after confirming it was just another fear-inducer service for people who get off on that.

the one useful thing Nextdoor is good for is lost/found pets. Other than that, i've never liked it.

I have a Ring and some of it is just downright funny.

> "suspicious person on my porch, be on the look out!" >> "um, that's the mailman."

I think the key term is "consumer-centric". Ring and other home cams are necessary peace of mind for a lot of people.

Consumers can be assholes as well. I was harassed by one of my neighbors and the police because their ring camera caught me walking through the parking lot in front of my building on the same afternoon this neighbors car was broken into.

Non-DIY plus subscriptions plus corporations rarely leads to "consumer-centric".

More often it's almost solely profit-driven.

One could argue that for most non-engineers, non-DIY is crucial for most products.

For example, in the 90s a lot of people thought everyone would host their own email servers. In reality, consumers flocked to Y! mail and later Gmail, since most people really can't or won't do this themselves.

Most people also do not change the oil on their own car (at least in the USA) etc.

In my mind the question is can you develop a great product in this area. It is possible (and sad) that Ring or Nest is as good as it gets (or have default won due to superior distribution). But I do think there are a lot of features I would want as a consumer that would make this experience better for me.

What's wrong with subscriptions? If you are getting an ongoing service (like Spotify, or HBO, or magazine delivery, etc...) it seems that an ongoing fee makes sense.

Products turning into services, that's what wrong. You stop paying, the devices stop working. Or, they decide they're bored with providing the service, and pivot or get acquihired (or just decide to prod you into upgrading), the devices stop working.

Plus, it seems that the new breed of service companies isn't satisfied with just providing a service in exchange of money. They also use the opportunity to exfiltrate as much data about you as they can get.

That is fine. I am happily paying for Spotify as they provide me with tremendous value day over day. I spend much less on music than in the days of buying CDs and Have access to a huge catalogue, instantly.

You don't subscribe to cable tv or netflix?

I do, but a) problems of these businesses are well-known, and b) there's a difference between a subscription to entertainment and subscription to necessities (or at least things with higher impact on one's life).

Also, in case of cable/Netflix, a lot of downsides are mitigated by piracy. There's no worry you'll be cut off from the entertainment you want, because you can always Torrent it if push comes to shove.

Eventually it's going to come to a head that these cams do little at all for security. I have a wyze cam that I use to watch my pet, but I'm well aware that even if it caught a burglar in the act that the LAPD isn't going to do anything with that video or the case.

I agree with you, but I think that particular example says more about our police than it does about the cameras.

About new social networks, I am building one that I am calling "a quiet social network".

Copy/pasting from my landing what it means:

A personal journal and a social network that will provide a quiet space to reflect about yourself and also to nurture your long term relationships with the people that you care about.

Why “quiet”?

Because this social network won't have the frenetic rhythm of news and updates of all other social networks.

A quiet space is just as quiet as the quietest sound. Quid Sentio is being designed so you will only listen to your voice and the voice of your close ones. A digital space to cultivate conversations more meaningful than the loud noise of social media and more long-lasting than the unsearchable small talk of instant messengers.

Quid Sentio is for you if you want to...

Avoid the deafening rumble of the crowds. You will only see public entries from people that have both being included in your list and included you in their list. No stranges following you (or even knowing you have an account).

Avoid the tiresome grumble of the acquaintances. There will be no way to search for people on the site or to see a list of your friends' friends. Also, there will be a limit of people you can add to your list without reciprocity, so no way to spam everyone in your contact list.

Avoid the popularity contests of perfect lives. There will be no way to like entries, only conversations. Also, there will be no way to share content outside of the list of who posted. So no way to go viral and no instant rewards.

Avoid the manipulative tricks of addiction dealers. All the design decisions above already point to a social network with less activity. Adding to that, you won't see any advertising on the site, so there is no incentive to keep you aimless wandering around here. Stay as long as you need to connect with your family, your close friends, and yourself. Then leave.

Avoid the news, the memes, the FOMO. As the posts all follow the same design of journal entries, with no images and no special treatment for external links, you will probably won't see much news or click-baity articles, unless a close friend wants to comment on them. Also, no company accounts either.

If you are interested: https://www.quidsentio.com

Your web site isn't loading at all for me. Glancing at the console it looks like you have a timeout error, and the loading gif just runs forever.

Chrome Version 79.0.3945.88 (Official Build) (64-bit)

Thanks! If it is the same bug that I experienced, if you refresh the page it will probably work.

I am a 3-week vacation right now, away from my development laptop, but I will investigate this bug better and solve it once I'm back

"The tab has crashed" after clicking on the "new entry" button.

Apparently it is not ready for primetime yet :( Thanks for letting me know!

I can't say as I like his take on the longevity industry. It is the take that will produce few meaningful advances, the "looking under the lamp because that's where the light is" way of approaching life. Just more marginally better drugs that do a little bit more than those of 10 years ago.

Sadly investors probably care very little from a financial position as to whether a drug works or not, as their exit usually happens somewhere between trials at Phase 1 and Phase 2. Earlier in the longevity market because it is hot.

I've put together Request for Startup lists for the longevity industry for the past few years, based on fairly detailed insight into the state of the science.




Because things move slowly in biotech, just about everything in these documents except for more senolytics is still valid.

Love your blog! Recently got interested in longevity research, as I was surprised how much closer to reality it is than I imagined, and also as a possible future industry to work in, and your blog has been one of the best resources to both get up to speed on the topic and also keep up to date!

As you are as familiar as few people are with the topic, I'd like to ask you for some advice, if possible: What do you think is the best way for someone like me (currently biochem undergrad, with ~7 years broad software engineering experience), to have an impact in the field?

Any bio* field can lead to working on treating aging; at undergraduate level you have tremendous flexibility as to which direction you take. For now build connections. Go to the conferences where industry meets science and meet people (Undoing Aging, Ending Age-Related Diseases, Longevity Therapeutics, Longevity Leaders, etc). Figure out who the people are you'd like to work with. Either for the corporate path of interning with companies working on aging, leading to a scientific position with one such company, or for the academic path of postgraduate work with a research who is doing something in aging that you find interesting.

Thanks for the response!

Dude, you are on it! Congratulations.

(FWIW, I think that space colonization will only be practical if we develop either FTL or longevity (live to 1000+) and the latter seems way more doable than the former, eh?)

I fail to see how TikTok is a "good reminder for the generational turnover of social products". Has an entire generation passed since Vine or Snapchat? If anything it's proof that securing an early lead in a new form of media (short form video in the aforementioned cases) doesn't guarantee success even for a generation. My takeaway is that young people are increasingly mercurial and disloyal. Chasing their attention seems like a recipe for disappointment.

The lifecycle of a social media should be measured by some % of pairs of children and their parents both using it. It’s an instant buzzkill.

And businesses.

This rule still holds true today. When I talk to people in their 20s and younger these days about Facebook, for instance, the near universal reaction is "Facebook is for old people and companies". They're all using something else.

Snapchat is still extremely popular among people my age (20s). Vine only died because it was killed off, the community was thriving. If anything, tiktok shows that it was a mistake to kill off vine; functionally it's the same sort of content.

i'm in my 20s snapchat is dead

Purple Air hits on a bit of the neighborhood pollution sensor idea: https://www2.purpleair.com/

I've used their API to look at California wildfire data.

Their API is great. I found their site a little slow for something I want to check regularly so I built a faster one based on the API:


If there's a PurpleAir sensor near you, it will show you the reading instantly. It also updates the favicon so you can leave it open in the background and check the reading just by glancing at the tab.

This is great! saved to home screen on iOS for convenience

Purple Air has been extremely popular in our community due to the wild fires here in California. There are quite a few sensors around town, and it gives you a good idea about how to prepare the kids for school and whether we should be riding bicycles.

How do you act on the air quality info? Seems to me that riding a bike or driving a car you'd be breathing the same air.

When you ride a bike you breathe a lot more. Staying indoors and exercising less (therefore breathing less) while you are outside are the two main things you can do when air quality is bad.

Light poles, cell towers, smart speakers and home internet routers seem like a good place where to bundle sensors with.

p.s. thanks for that site. Works well with BitBar:

      /usr/bin/sudo -u dzhiurgis -i bash -c 'echo -n "AQI "; curl -s https://www.purpleair.com/json\?key\=EAOGA5Q4JOPE8HH7\&show\=17325 | jq -r ".results[0].PM2_5Value"'

I think this data is integrated into weather underground app as well.

> What would be a network which allowed for more thoughtful discourse

We used to have this and it was called forum (phpbb and such). They have been oblitared and we moved to Facebook apparently leaving our brains behind.

In reality we are victims of armies of psychologists optimizing for engagement.

I have the feeling forums will come back though.

We didn't leave our brains behind, this is just the internet without heavy handed moderators keeping subforums organized, discussions on topic, and putting the brakes on flame wars. Of course things crash and burn in an environment of total anarchy like what is seen on social media websites, image boards have taught us that decades ago.

Yes, we didn't leave our brains behind. The environment in modern social media rewards the wrong type of interaction.

Also, it's the internet now that everyone is on it. Heavy moderation of a niche group is pretty easy. Try moderating 100 Million + active accounts.

I would say reddit is a forum, and you can have good discourse on there, but can also have straight comfort internet trash. Basically I think its more nuanced than just having a forum...

Reddit is probably the closest thing to a forum. I guess having a separate domain in a website owned by people rather than one run by a big company allows for a more intimate environment.

I don't really find reddit comparable to forums, mainly because reddit is also built around recent comments. It has very rudimentary search functionality, and the way submissions and comments are scored discourages long conversations. Unless you're on a really small subreddit, stuff falls off the first page quickly, and your comment will be far less visible if posted half a day later. This is completely different than forums, which optimize for conversations that last for several days to a couple of weeks, sometimes much longer.

Exactly. Forums optimize for quality content and this makes it hard to monetize them because when you found the good stuff you are not going to click on ads.

Also, mobile which is not ideal for long conversations helped their demise.

Still, I hope they come back big time. I miss them.

The forums I hung out on simply sorted by the last comment, or by votes. They certainly didn't "optimize for quality content" in any particular way, other than banning trolls (sometimes.)

Also plenty of good forums had ads. I don't think it's a common thing to want to refuse to click on ads because of high quality content, I would guess the opposite tends to be true.

I probably have a too much romantic memory of them :)

Anyway, it was easy to come back to a specific topic, so it was easier to find good ones. You could bookmark them, see the number of views/replies or even find them in search engines.

Good content surfaced and survived better, so optimization is probably the wrong word.

#1. Trinet for full-time remote/distributed workers. I've spent countless hours researching Global PEOs and reviewing their often poorly written contracts. If a reputable service existed that allowed us to hire in any country without having to form our own corporate entity in that country, we'd be ready to sign yesterday. Unfortunately, all existing services are either:

* Too new (There are promising upstarts, but they usually don't operate their own entities and it seems risky to route all our IP ownership assignments through a tiny company)

* Too expensive (massive markups on what should be a standardized service)

* Too incompetent (One PEO sent us a contract for a Canadian employee that assigned their IP in accordance with US law. It's facepalm-bad sometimes).

What does PEO mean and what existing services have you found?

PEO stands for "Professional Employer Organization" and is an American term for a company that serves as the legal employer for the people you want to hire. For example, see https://justworks.com/. When an employee joins our company, they legally become an employee of JustWorks, but we're the practical employers. It's purely a legal relationship.

This is welcome the USA government because they know these companies help us achieve compliance better than we could on our own. JustWorks has been great, but only employs people in the USA. When we want to employ someone outside the USA, we therefore need to find a "global PEO." These companies are sometimes called "Employers of Record."

Popular Global PEOs are:

- Globalization Partners

- Elements Global

- Capital GES

- Pilot.co

- Lots more.

But it's a highly fragmented market and has been a pain to engage.

While only a footnote, I like the nod to nuclear energy.

I remember in the 1950s, there was a Popular Mechanics cover touting nuclear as a coming technology to power homes and even cars.

Imagine the next Tesla-like company offering to install a small nuclear reactor in houses. Like solar, you can sell electricity back to the grid, charge your electric car with it, literally use it to heat your water....

Also might be a very attractive option to going off-grid.

Nuclear reactors don’t scale down very well. The economics really break down under about 10MW which is ~5,000 times more power than most homes need.

Even ultra cost efficient full scale powerplants are a money losing proposition without huge subsidies. They are being squeezed between extremely cheap renewables, low cost natural gas, and rapidly advancing storage technology.

Which is why nuclear was 17.6% of global electricity generation in the late 90’s but has fallen to 10% in 2019.

>Nuclear reactors don’t scale down very well. The economics really break down under about 10MW which is ~5,000 times more power than most homes need.

This is true and fully agreed. Home reactors are not really feasible or even a good idea.

>Even ultra cost efficient full scale powerplants are a money losing proposition without huge subsidies. They are being squeezed between extremely cheap renewables, low cost natural gas, and rapidly advancing storage technology.

Plants are expensive due to the regulation and insurance making them that way. Safety with nuclear is obviously paramount, but political and social pressure has expanded a lot of the requirements to quite a high degree. Additionally, renewable are also subsidized quite a lot and are not always viable in all locales. Storage tech is good for all forms of power generation.

>Which is why nuclear was 17.6% of global electricity generation in the late 90’s but has fallen to 10% in 2019.

Nuclear also has a lot of fearmongering and red tape around it. Plants are reaching their EOL and in many cases new ones are not being built due to the cost of getting through all the red tape (getting approval from the feds down to local government, dealing with inevitable NIMBY lawsuits, etc). Renewables are the future, but we aren't fully in the future yet.

> Nuclear also has a lot of fearmongering and red tape around it. Plants are reaching their EOL and in many cases new ones are not being built due to the cost of getting through all the red tape (getting approval from the feds down to local government, dealing with inevitable NIMBY lawsuits, etc). Renewables are the future, but we aren't fully in the future yet.

This is a common trope.

I can't work out where you live, but I do often see this line of reasoning put forward by residents of the USA, who overlook the fact that 'the feds' don't regulate what happens in the other 95% of the planet, and yet other parts of the world are equally disinterested in building more nuclear fission plants.

It may be that the problems -- federal / national governments red tape, NIMBY lawsuits, local gov / state gov / provinces / councils, 'fearmongering', etc -- are common and similar everywhere, but this seems prima facie unlikely, certainly not demonstrated.


Ontario sits at about 2/3 nuclear. We have for a long time. And our electricity is relatively cheap.

I was going to point out that Ontario, at 15m, accounts for 0.002% of the population, and suggest it may be a statistical blip.

But -- and thanks for the link, it's fascinating -- diving into that page raises some questions. The nuclear component is very steady at 11MW, which I suppose is a function of the technology, to provide a reliable base load.

Nuclear fission accounts for 12MW -- while Hydro seems to consistently provide about a third that figure, with wind averaging just over 1MW, but peaking up to 4MW.

Canada has 19 nuclear fission power plants, with all but one being in Ontario.

Seemingly [1] several were / will be shutdown, as refurbishment is too expensive, and no plans to construct new plants in the province.

Storage of spent fuel is expected to cost ~$30B ... but that project hasn't found a site yet, and that's a projected figure circa 2005.

[1] https://en.wikipedia.org/wiki/Nuclear_power_in_Canada

The import vs export numbers are critical for understanding what’s going on. The reactors are located in Ontario, but they supply power across a much wider area. Further, abundant hydroelectric power allows them to load follow locally at low cost. Yet they are still often paying 15+c/kWh.

The hydro power is actually pretty fascinating in what it might mean for usage of unstable renewables like solar and wind. Hydro is a giant, slowly recharging battery. It's perfect to pair with wind and solar because it can deal with their highly variable generation and our society's highly variable needs.

But we don't have enough hydro-battery storage to allow for solar and wind to provide our base load. That's where nuclear shines, and why Ontario uses it so much.

Until we can come up with a way to store something like a week's worth of power, it's going to be very difficult to ever phase out of nuclear. But given how clean nuclear is compared to the alternative base-load energy sources (gas, coal, etc), I think that's just fine.

It might seem that way, but minimum and maximum power demand are generally so far apart when you include seasonal differences, solar and nuclear need about as much peaking power to supplement each other. That’s why even with huge exports Frances nuclear power stations hover around 77% utilization where US nuclear can hit 92%. Nuclear can clearly play a role, especially near the poles, but it’s just really expensive.

In the end Ontario is in an unusual situation that stops working as the rest of Canada scales either nuclear or wind power. It’s working for them and their electricity is mostly carbon neutral, it just does not scale.

PS: That’s not to say nuclear can’t scale to 100%, nuclear subs make it work with wildly swinging demand. However, low capacity factor directly results in higher costs.

Thanks mabbo, Retric.

Given 10-15 years ago there were very definite plans to build new plants, but these [1] all appear to have lapsed or been deferred indefinitely, this further supports my original point -- even seemingly enlightened & pro-nuclear societies / governments are pulling back on their nuclear fission build-out plans.

[1] https://www.world-nuclear.org/information-library/country-pr...

But there are places with much larger portions of their power supplied from nuclear plants, e.g. France, with >70% nuclear power.

That’s incorrect.

France generation is over 70% nuclear power, but in 2018 exported 86.3 TWh mostly nuclear and imported 26TWh mostly non nuclear. So they actually use less than 70% nuclear power. This is important because of how hard it is to match nuclear generation with demand resulting in a 77% capacity factor vs ~90% in the US nuclear industry even with all those exports.

In other words they built enough nuclear to cover well over 80% of demand, but only use ~60% internally.

They are also looking to significantly reduce nuclear production. Closing up to 17 reactors by 2025. https://www.france24.com/en/20170710-france-hulot-could-clos...

Yup, undeniably France is an interesting outlier.

As I understand it, the bulk of their reactors were produced over a relatively short boom period of construction, and their projected EOL's were consequently going to coincide -- the number of safety incidents, ageing designs, costs of attending to both, public opinion swaying against, resulted in the relatively recent decisions to try to reduce the reliance on fission.

But, even so, this speaks to my point -- in countries with a positive history of nuclear fission, and little 'red tape' to stall new builds, they're still moving away from it.

I worked in the nuclear power in the past. The massive volumes of red tape involved is not conjecture, I experienced it firsthand. Activists will stonewall any new construction, full stop, and the NRC requirements for existing plants only go up over time.

> I worked in the nuclear power in the past.


My point was and remains that USA federal government hoops around building out new nuclear fission plants can not explain the dearth of new nuclear fission plants being built everywhere else.

Canada was cited elsewhere in this thread as pro nuclear fission plants -- but they've deferred all plans to build new plants.

France is the really obvious big proponent, but their recent experiences (building Flamanville in FR, involvement in Hinkley in the UK) suggest there is more complexity here than can be addressed by complaints that 'The USA government has red tape'.

Nuclear professional here. It is true that economics for electricity-only plants have historically struggled to compete with conventional sources, actually below closer to 800 MWe.

The Small Modular Reactor hypothesis is that economies of mass production could conceptually be built that overpower economies of scale. The small reactors of the past, including the ML-1 truck-mounted military microreactor were 10x too expensive even for the military in remote areas. Just building a few small reactors is a losing proposition. If you can get them to "go viral" before they've achieved economic parity, then there's a chance. That will only happen if you successfully market their 24/7 very-low-carbon, very-low land, very-low raw material footprints.

I think for climage change purposes, we should focus on getting costs down on 500-1000 MWe plants. If it takes another round of small prototype non-LWR reactors to get there, then so be it. But large stations are what will displace most of the 84% of the world's energy that is fossil fueled.

Regarding the competition, low-cost fracked natural gas has been deadly to nuclear. I don't think most people realize that fracked natural gas, while great in the deadly air pollution department, is just as bad as coal in the climate change department (when you factor in the methane leaks from wells and pipelines). So if markets can price carbon emissions, natural gas can be ruled out. If not, natural gas will continue to drive nuclear plants to closure and then replace them.

Extremely cheap renewables are a friendly competition to nuclear in that if they're successful, the goals of the nuclear proponents are met: clean, plentiful, cheap energy 24/7. At the moment, the major issues of land use, raw material use, and intermittency are not causing much trouble for renewables. But as they scale up they may encounter more difficulties. All energy sources experience new troubles and regulations as they scale. Coal got scrubbers and filters (doubling+ capital costs), nuclear got the NRC, solar in California recently ran into NIMBY in San Bernardino county desert. Will that continue to get worse? Or are the positive attributes of renewables so good that people will continue to embrace at scale? I honestly don't know. I keep working on nuclear because it's a good high-density resource.

TL;DR: Include carbon-free as valuable in markets and nuclear would do great.

Yes, absolutely.

Today’s modern reactors are night and day compared to Chernobyl/Fukushima era reactors - they’re not even comparable. They’re fail-safe rather than fail-deadly, and are much more compact and efficient, with better controls and containment. The size of the reaction chamber is basically that of a household washing machine.

I’d gladly live next door to a modern plant.

Current reactors are extremely safe compared to fossil if you look at the stats. 3 orders of magnitude safer in a deaths/TWh basis. The new designs (which are actually all revivals of 1960 designs) may be safer and cheaper but we have yet to prove that.

I'm in the middle of writing an elaborate history of the quest for economical nuclear power to help with this discussion.

Sorry for being pedantic, but do you mean fail-secure? Fail-safe = things inside the area can get out on failure. Fail-secure = things cannot get out i.e. catastrophe is contained (though people inside may die).

I don't know much about reactor design, but I don't think I'd live next door to a fail-safe plant, but please correct me if I'm misguided because.

In general, fossil kills 4-6 million people per year from air pollution. Nuclear has killed ~4000 total, ever. So nuclear plants are very, very safe compared to normal energy alternatives.

Fail-safe just means that if equipment breaks or a human does something wrong, the plant goes to low power and passive natural-circulation systems kick-in that keep the low-power shutdown mode from going to temperatures high enough to break the radiation containment structures.

What exactly do you mean by fail safe? Are there no catastrophic failure modes? So even if a group of people with malicious intent were to gain access, they couldn't do more damage than with a natural gas or coal burning plant?

Not parent commenter, but I'll repeat what I said above to someone else about what fail-safe means in terms of nuclear reactors:

Fail-safe just means that if equipment breaks or a human does something wrong, the plant goes to low power and passive natural-circulation systems kick-in that keep the low-power shutdown mode from going to temperatures high enough to break the radiation containment structures.

Recall that coal plants and oil emissions kill 4-6 million people per year from air pollution. Nuclear plants are crazy safe compared to that. And while natural gas is safe from an air pollution POV, the hazards of climate change from it are potentially large. Nuclear reactors are essentially carbon-free.

I dream of the day when all the oil derricks and refineries hidden behind facades all over west Los Angeles would be replaced by nuclear energy, not to mention the inglewood oil field.

I’m pretty sure my HOA has a clause against installing a personal nuclear reactor in our home :)

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact